AMD markets their RX 400 series as the best for 1080p gaming

[WYP]

AMD markets their RX 400 series as the best for 1080p gaming, releasing benchmark results comparing the RX 470 to the unreleased GTX 1050 and 1050 Ti.



Read more on AMD's pre-release GTX 1050Ti benchmarks.

 

[Greenback]

Well if you want to be honest the 1060 is better for 1080p if you consider that DX12 isn't in full use atm, so most games are DX11 and the 1060 is better mostly. By the time DX12 becomes the mainstream it will probably be time to upgrade the GPU anyway

 

[NeverBackDown]

The 1060 isn't relevant in this comparison though, for budget 1080p the 470 and 1050ti are competing. That's the point they are trying to make, even if they didn't do it well

 

[semitope]

Well if you want to be honest the 1060 is better for 1080p if you consider that DX12 isn't in full use atm, so most games are DX11 and the 1060 is better mostly. By the time DX12 becomes the mainstream it will probably be time to upgrade the GPU anyway

dx12 became mainstream after the last AAA game to use it. Its only going to become more prevalent in 2017.

 

[NeverBackDown]

That's not what mainstream is.. once it becomes a regular thing and many AAA titles support it, then yes mainstream. 2017 should be a much bigger dx12 year though, that I will agree with

 

[Kushiro]

Now how did you get your hands on that AMD.

Looks like they forgot to let people know that free sync comes at a premium too. Then again it is royalty free

 

[Kilbane]

Am I the only one that is confused by the way AMD try to advertise their products?

I am going to pick on one slide here from the presentation (as I can happily pick holes in all of them). Specifically this one :

Starting from the top, Kudos to AMD. These results are exactly the same as our own testing on BF1 with the 480 vs 1060 using DX12 (And I mean EXACTLY the same as the results in the BF1 performance test).

Now there is no denying that that top graph holds a lot of merit. This is where they should have left it.

Moving on to the 470 vs 1050 Ti and the 460 vs ....... Hold on, a 750 Ti. Yes a 750 Ti, a GPU that has been replaced twice in generations. I don't understand that to start with, why not compare to the 1050, maybe even a 950?

So this got me thinking that maybe they compared to price points. So the 470 costs £190 4GB and £220 8GB (Prices accurate from Scan at time of writing this reply), the 1050 Ti however is set to release at £135 (Unofficial but the best answer I could find before the release).

Understandably, the line up of AMD vs nvidia does somewhat place these cards together. But if you consider that the 470 costs a minimum of £55 more than the 1050 Ti, would it not be fair to say that if you payed that over the 470 you in fact start in the territory of the 1060.

OK, I'm grasping at straws there. But here is the really interesting part. For a 4GB RX 460, you will pay £125. The same price as a 1050 Ti, which according to AMD would be the better card to buy for the 'Biggest game of 2016'.

So, in that case. The 460 is in fact the card to be placed against the 1050 Ti, meaning that the 470 is actually in no mans land between the 1050 Ti and the 1060. Which should be a selling point on its own for AMD, Bang for Buck, the 470 is in a great place for those who can't stretch to the budget of the 1060/480 but want something above the 460/1050 Ti.

TLDR; AMD need to stop making pointless charts that always show them on top by dragging old cards into the mix. Just advertise the unique selling points such as DX12 performance of the 480 and the price to performance ratio of the 470. Forget the 460, if the price leaks are correct for the 1050 Ti, no one will consider the 460 after that.

 

[NeverBackDown]

While you can nitpick AMDs marketing all the time nearly everytime, you have to understand Nvidia are just as bad, notoriously bad unequal charts(ie, make a 5% look more like a 30%). Most companies do it like this, just AMD is terrible at it. So really everyone needs to improve, but I understand your point, AMD needs to get better at it instead of always taking the same angle.

 

[S_I_N]

I'm sorry but doing comparisons by price of unit imho is shady. I think it should side by side like 480 to 1080 470 to 1070 and so on comparing their 2nd card to nvidias lowest imho is stacking the results in their favor.

 

[Kushiro]

I'm sorry but doing comparisons by price of unit imho is shady. I think it should side by side like 480 to 1080 470 to 1070 and so on comparing their 2nd card to nvidias lowest imho is stacking the results in their favor.

Did you even read what you are posting exactly?

The comparisons you are asking for arent even in the same price range let alone performance.. Both nVidia and AMDs lineup is not complete. AMD doesnt really have a GPU for the high end yet. The RX 480 is mid range.

 

[S_I_N]

Did you even read what you are posting exactly?

The comparisons you are asking for arent even in the same price range let alone performance.. Both nVidia and AMDs lineup is not complete. AMD doesnt really have a GPU for the high end yet. The RX 480 is mid range.

Yes I did know exactly what I posted did you even bother reading it??? I said comparing by cost of unit was imho shady instead compare along the same line why compare the AMD second card from top to nvidias lowest card imho it stacks the results. I know the cost are vastly different cause AMD has always been a lower cost option

 

[Davva2004]

You have to look at price vs performance as a metric. If you don't, then you can say that the Titan XP beats the RX480 in every game. Of course it does. You're comparing Porsches to Fiestas and saying the Porsche is always the better car because it's faster!

 

[SeekaX]

Can manufacturers please stop advertising their products with bnenchmarks vs their competitor's products? They have absolutely no value as long as the people who release the benchmarks are biased.

 

[S_I_N]

You have to look at price vs performance as a metric. If you don't, then you can say that the Titan XP beats the RX480 in every game. Of course it does. You're comparing Porsches to Fiestas and saying the Porsche is always the better car because it's faster!

I don't base off price when it comes to gpus cause well nvidia always seem to charge higher dollars ( and gets it) than AMD does but I still think comparison should be straight across the lines not here's our second highest offering were gonna compare it to your lowest cause it's priced roughly the same just how I see it. Comparing a Porsche to a fiesta is laughable. We all know the fiesta wins hands down

 

[Kushiro]

Yes I did know exactly what I posted did you even bother reading it??? I said comparing by cost of unit was imho shady instead compare along the same line why compare the AMD second card from top to nvidias lowest card imho it stacks the results. I know the cost are vastly different cause AMD has always been a lower cost option

Yes I did bother to read your post. Shame you couldn't do the same.

You cannot compare an RX 480 to a 1080 , let alone a Titan Xp end of story. If AMD marketed it as a high end GPU and prices it the same as a 1080 then by all means. But you are literally asking for them to compare a card that is in general twice the price. While we are at it, should we cut benchmark scores in half just cause it costs more?

 

[S_I_N]

Look I read it all and while I see the merit of what it's about I myself find it somewhat stacked I understand the cost factor involved in the comparison I disagree with comparing a higher end gpu to the lowest gpu even if costs are close. It's my opinion and I am entitled to it. If the costs were aligned between the 480 and 1080 this would all be moot as nvidia asks for more $$ and AMD Doesn't it's a buyers market. I'm simply saying is DUH!!! A 480 beats a 1050ti that's a no brainer

 

[NeverBackDown]

Look I read it all and while I see the merit of what it's about I myself find it somewhat stacked I understand the cost factor involved in the comparison I disagree with comparing a higher end gpu to the lowest gpu even if costs are close. It's my opinion and I am entitled to it. If the costs were aligned between the 480 and 1080 this would all be moot as nvidia asks for more $$ and AMD Doesn't it's a buyers market. I'm simply saying is DUH!!! A 480 beats a 1050ti that's a no brainer

I'm trying to find your argument here and how you see things, but honestly I can't see it. It might be bad wording or just me personally overlooking it. But I can see why Kushiro is debating here.

As far as I can tell, you are talking about comparing straight across the line here as in , 1080 vs 480 because if the "80" in both. So they should be equal in comparison. However I think that's not fair because rarely do both companies line up there naming schemes.
You also have take into account architecture and design. Pascal GP102(or whatever the heck 1080 is) is a much bigger die and high performance design than a 480 is. A 480 is a smaller die and designed for more power efficiency per transistor(not saying it's more efficient, but by AMDs standards it is). So they aren't at the same level here and a comparison is just naïve. A 1060 and a 480 are the closest comparison(that's fair because of reasons above) that we have. So anything below that, 470 vs 1050ti are the next closest. Technically, both are the 2nd tier here and a comparison is relatively close.

I think if you are trying to go by what you're argument is(if I understood it correctly) a better and more accurate way of doing so is die size and transistor count(meaning transistor/^2 nm). This imo is probably a better argument than using a straight across the line method you were using. Just food for thought